Seismic Behavior of Precast Hollow-Core FRP-Concrete-Steel

Column having Socket Connection Mohanad M. Abdulazeez1, Mohamed A. ElGawady2 1 Graduate Research Assistance and PhD student, Dept. of Civil, Architectural, and Environmental

Engineering, Missouri University of Science and Technology, Rolla, MO. 65409 2 Benavides Associate Professor, Dept. of Civil, Architectural, and Environmental Engineering, Missouri

University of Science and Technology, Rolla, MO. 65409

ABSTRACT

This paper experimentally investigates the seismic behavior of a large-scale hollow-core fiber-

reinforced polymer-concrete-steel HC-FCS column under cyclic loading. The typical precast

HC-FCS member consists of a concrete wall sandwiched between an outer fiber-reinforced

polymer (FRP) tube and an inner steel tube. The FRP tube provides continuous confinement for

the concrete wall along the height of the column. The column was inserted into the footing in a

socket connection with an embedded steel tube length of 25 inches and temporary supported.

Then, the footing was cast in place around the column. The seismic performance of the precast

HC-FCS columns was compared and assessed with previous experimental work. The compared

column had the same geometric properties but the thickness of the steel tube was 25% thicker

than the tested column in this study. This study revealed that the assembly of these HC-FCS

columns was deemed satisfactory by developing the whole performance of such columns and

providing excellent ductility with inelastic deformation capacity by alleviating the damage at

high lateral drift.

INTRODUCTION

The Federal National Bridge Inventory (FHWA 2013) classifies 63,522 bridges as “structurally

deficient” and 84,348 bridges as ”functionally obsolete” with many of them needing to be

repaired, rehabilitated, or replaced. Therefore, a rapid construction method to address this

challenge is highly needed.

Accelerating bridge construction (ABC) will reduce traffic disruptions and life-cycle costs as

well as improve construction quality and safety, resulting in more sustainable development

(Dawood et al. 2014). The use of precast concrete bridge elements is one strategy that can reduce

on-site construction time, field labor requirements and traffic impact. Precasting also improves

the safety and quality of construction.

Recently, interest has been rapidly growing in using fiber reinforced polymer (FRP) tubes in

construction as a replacement for the outer steel tube of the DSTCs column (Teng and Lam 2004;

Teng et al. 2007). The proposed column is introduced as hollow core FRP-Concrete-Steel (HC-

FCS) columns. The column consists of an inner steel tube and an outer FRP tube, with a concrete

shell placed in-between the two tubes (Fig. 1). FRP fibers are oriented in the hoop direction to

increase the concrete confinement and provide more shear resistance (Zhang et al. 2012). HC-

FCS composite columns have several advantages as a precast element over conventional

reinforced concrete or structural steel and can be delivered from the combination of all three-

component materials. The concrete infill is confined by both FRP and steel tubes, which results

in triaxial state of compression that increases the strength and strain capacity of the concrete infill

and enhances the seismic performance of such columns (Abdelkarim and ElGawady 2014;

Abdelkarim and ElGawady 2015; Abdelkarim and ElGawady 2016; Abdulazeez et al. 2017).

The main objective of this study is to investigate the performance of HC-FCS columns under

axial and static dynamic loads. Hence, achieving robust precast column under inelastic cyclic

deformations easy to construct with high quality.

(a) (b)

Figure 1. The general arrangement of the HC-FCS column (a) inserting the steel tube and casting the

footing (cast-in-place socket connection); (b) installing the FRP tube and pouring the concrete wall.

EXPERIMENTAL PROGRAM

In this study, a 0.4-scale HC-FCS column, F4-24-E3(1.5)4, with an embedment steel tube 635

mm in length corresponding to 1.6 Di (Di is the inner diameter of the steel tube) was tested under

constant axial load and lateral cyclic load. The F4-24-E3(1.5)4 column had a circular cross

section with an outer diameter of 610 mm and a clear height of 2,032 mm. The lateral load was

applied at a height of 2,413 mm with a shear span-to-depth ratio of approximately 4.0. The

column consisted of an outer filament-wound GFRP tube with a constant thickness of 9.5 mm

along the height of the column. The inner steel tube had an outer diameter of 406 mm and a

thickness of 4.8 mm. A concrete wall with a thickness of 102 mm was sandwiched between the

steel and FRP tubes (Fig. 2).

The columns’ label used in the current experimental work consisted of four segments. The first

segment is a letter F in reference to flexural testing, followed by the column’s height-to-outer

diameter ratio (H/Do). The second segment refers to the column’s outer diameter (Do) in inches.

The third segment refers to the GFRP matrix using “E” for epoxy base matrices; this is followed

by the GFRP thickness in 1/8 inch (3.2 mm), steel thickness in 1/8 inch (3.2 mm), and concrete

wall thickness in inches (25.4 mm).

The concrete footing that was used in this study had a length x width x depth of 1,524 mm x

1,220 mm x 864 mm with bottom reinforcements of 7-#7, top reinforcements of 6-#7, and shear

reinforcement of #4 at 64 mm (Fig. 2). The steel cage of the footing was installed into the

formwork. The construction steps are (1) preparing and installing the reinforcement cages of the

footings, (2) installing the steel tube inside the footing cage with an embedded length of 635 mm,

(3) pouring the concrete of the footing [Fig. 3 (a)], (4) installing the GFRP tube and pouring the

concrete of the column [Fig. 3 (b)], (5) installing the reinforcement cage of the column head and

concrete pouring [Fig. 3 (c)].

The mechanical properties of the steel tube and rebar are summarized in Table 1. The rebar

properties are based on the manufacturer’s data sheet, while the steel tube properties were

determined through tensile steel-coupon testing according to ASTM A 1067. The concrete mix

design is shown in Table 2. Pea gravel with a maximum aggregate size of 9.5 mm was used for

concrete mixtures. Table 3 summarizes the unconfined concrete strength for the footings and

columns. The material properties of the glass FRP tubes are presented in Table 4.

(a) (b)

Figure 2. Construction layout of the column (a) Elevation, (b) column cross-section.

(a)

(b)

(c) (d)

Figure 3. Constructing procedure (a) footing casting around the inner steel tube; (b) placing the FRP

tube and pouring the concrete wall; (c) column head casting; (d) the HC-FCS column.

Table 1. Steel properties of the rebars and steel tube

Elastic modulus

(E, GPa)

Yield stress

(fy, MPa)

Ultimate stress

(fu, MPa)

Ultimate strain

(ɛu, in/in)

Steel rebar 200 413.7 620.5 0.08

Steel tube 200 399 441 0.22

Table 2. Concrete mixture proportions

w/c Cement

(kg/m3)

Fly Ash

(kg/m3)

Water

(kg/m3)

Fine aggregate

(kg/m3)

Coarse aggregate

(kg/m3)

0.5 350 101 225 848 848

Table 3. Summary of the unconfined concrete strength of the columns and the footings

F4-24-E3(1.5)4 F4-24-P324

Column Footing Column Footing

f’c at 28 days (MPa) 35 55 32.6 36.6

f’c day of the test (MPa) 46.5 56.7 36 39

Table 4. FRP tubes properties

Elastic modulus

(GPa)

Hoop elastic Modulus

(GPa)

Axial ultimate stress

(MPa)

Hoop rupture stress

(MPa)

4.7 21 83.8 276.8

EXPERIMENTAL SET-UP AND INSTRUMENTATIONS

Sixteen linear variable displacement transducers (LVDTs) and string potentiometers (SPs) were

assigned for measuring displacement along the tested column F4-24-E3(1.5)4. A layout of the

LDVTs and SPs is shown in Figure 5 (a). Four LVDTs were mounted on each of the north and

south faces for the vertical displacement measurements at the potential plastic hinge region. Two

more LVDTs were attached for measuring the uplift and sliding of the footing during the test.

The effects of the footing uplift and sliding were considered before calculating the lateral

displacement and curvature of the column.

Forty strain gauges were installed on the FRP tube at five levels with 127 mm spacing between

them. Four horizontal and four vertical strain gauges were installed at each level, as shown in

Figure 5 (b). Fifty-six strain gauges were installed inside the steel tube at seven levels with a

spacing of 127 mm [Fig. 5 (c)]. Four horizontal and four vertical strain gauges were installed at

each level. A high definition webcam was placed inside the steel tube 635 mm from the top of

the footing level to broadcast buckling of the steel tube and any internal damage.

(a) (b) (c)

Figure 5. Strain gauges layout: (a) LVDT’s and SP’s installing; (b) mounted on GFRP tube; and (c)

mounted on steel tube.

LOADING PROTOCOL AND TEST SETUP

Constant axial load, P, of 489.3 kN corresponding to 5% of the axial load capacity of the

equivalent RC-column, Po, with the same diameter and 1% longitudinal reinforcement ratio was

applied to the column using six external prestressing strands (Fig. 6). The Po was calculated using

Eq. 1 (AASHTO-LRFD 2012):

𝑃𝑜 = 𝐴𝑠 𝐹𝑦 + 0.85 (𝐴𝑐 − 𝐴𝑠) 𝑓′𝑐 (1)

where 𝐴𝑠 = the cross-sectional area of the longitudinal steel reinforcements, 𝐴𝑐 = the cross-

sectional area of the concrete column, 𝐹𝑦 = the yield stress of the longitudinal steel

reinforcements, and 𝑓′𝑐 = the cylindrical concrete’s unconfined compressive stress. The

prestressing strands were supported by a rigid steel beam atop the column and the column’s

footing. The prestressing force was applied using two servo-controlled jacks to keep the

prestressing force constant during the test.

(a) (b)

Figure 6. (a) Layout of the test setup; (b) HC-FCS column ready for testing.

North South

After applying the axial load, the cyclic lateral load was applied in a displacement control using

two hydraulic actuators connected to the column loading stub. The loading regime is based on

the recommendations of FEMA 2007, where the displacement amplitude ai+1 of the step i+1 is

1.4 times the displacement amplitude of the proceeding step (ai). Two cycles were executed for

each displacement amplitude. Figure 7 illustrates the loading regime of the cyclic lateral

displacement. Each loading cycle was applied in 100 sec. corresponding to a loading rate that

ranged from 0.254 mm/sec. to 1.27 mm/sec.

Figure 7. Lateral displacement loading regime. RESULTS AND DISCUSSIONS

The moment-lateral drift plot is shown in Figure 8 (a). The lateral drift (𝛿) was calculated by

dividing the lateral displacement measured from the actuators displacement transducers by the

shear span of 2,413 mm. The moment (M) at the base of the column was obtained by multiplying

the force measured by the actuators’ loading cells by the column’s height of 2,413 mm. Figure 8

(a) and Table 5 compares the cyclic response of the two columns, F4-24-E324, and F4-24-

E3(1.5)4. As shown in the figure and table, column F4-24-E3(1.5)4 displayed 7% lower flexural

strength of 680 kN.m and 35% lower maximum lateral drift of 8.4% compared to column F4-24-

E324. Gradual stiffness degradation occurred beyond that until the end of the test. The stiffness

degradation occurred due to concrete damage inside the tube; steel tube buckling led to the

initiation of ductile tearing at 8.0% [Fig. 8 (b)] at the column-footing interface area and thereby

ended the test.

Table 5. Summary of the results of both column

Tested Column

Moment

capacity

(kN.m)

Lateral drift

at max.

moment (%)

Lateral drift at

failure (%) Mode of failure

F4-24-E3(1.5)4 680.0 2.7 8.4

Steel tube local buckling,

concrete crushing, and steel

tube ductile tearing

F4-24-E324

(Abdelkarim et al. 2015) 732.0 2.8 13.0

Steel tube local buckling,

concrete crushing, and FRP

tube rupture

For the investigated columns, the energy dissipation at each lateral drift was determined as the

area enclosed in the hysteretic loop of the first cycle at this drift level. Dissipating higher

hysteretic energy reduces the seismic demand on a structure. Figure 9 illustrates the cumulative

energy dissipation vs. lateral drift relation for columns F4-24-E3(1.5)4 and F4-24-E324 that have

been tested by (Abdelkarim et al. 2015). As shown in the figure, both columns dissipated the

same level of energy until drift of approximately 2.5%. Beyond that, column F4-24-E3(1.5)4

dissipated much less energy due to the severe local buckling followed by the ductile tearing of

the steel tube. At drift of 8.4% when column F4-24-E3(1.5)4 failed, column F4-24-E324

dissipated 240% higher energy.

(a) (b)

Figure 8. Moment vs. lateral drift for columns F4-24-E3(1.5)4 and F4-24-E324 (Abdelkarim et al. 2015),

(b) ductile tearing of the steel tube at 8% lateral drift.

Figure 9. Cumulative energy dissipation vs. lateral drift for columns F4-24 E3(1.5)4 and F4-24-E324

(Abdelkarim et al. 2015).

CONCLUSIONS

This paper presents the experimental results of a hollow-core fiber reinforced polymer concrete

steel (HC-FCS) precast column. The HC-FCS column consists of a concrete hollow cylinder

sandwiched between an outer fiber-reinforced polymer (FRP) tube and an inner steel tube. The

column had an outer diameter of 610 mm, an inner steel tube diameter of 406 mm, and a height-

Steel tube ductile tearing

to-diameter ratio of 4.0. The steel tube was embedded into reinforced concrete footing with an

embedded length 1.6 times the steel tube diameter, while the FRP tube acted as a formwork and

provided a continuous confinement for the concrete wall and was curtailed at the top surface of

the footing. The column was subjected to constant axial load and lateral cyclic load during this

study and compared to the HC-FCS column that was tested by (Abdelkarim et al. 2015) under

the same loading regime. The HC-FCS F4-24-E3(1.5)4 column failed at the drift of 8.4% due to

concrete wall crashing and steel tube buckling followed by ductile tearing at the column-footing

interface level.

ACKNOWLEDGEMENT

Missouri University of Science and Technology conducted this research with funding provided

by Missouri Department of Transportation (MoDOT) and Mid-American Transportation Center

(MATC). Gratefully contribution from ATLAS Tube is appreciated. The discounts on FRP tubes

from Grace Composites and FRP Bridge Drain Pipe are also appreciated.

REFERENCES

Abdelkarim, O. I., and ElGawady, M. A. (2014). "Analytical and Finite-Element Modeling of

FRP-Concrete-Steel Double-Skin Tubular Columns." Journal of Bridge Engineering.

Abdelkarim, O. I., and ElGawady, M. A. (2015). "Concrete-Filled-Large Deformable FRP

Tubular Columns under Axial Compressive Loading." Fibers, 3(4), 432-449.

Abdelkarim, O. I., and ElGawady, M. A. (2016). "Behavior of hollow FRP–concrete–steel

columns under static cyclic axial compressive loading." Engineering Structures, 123,

77-88.

Abdelkarim, O. I., Gheni, A., Anumolu, S., and ElGawady, M. A. "Seismic behavior of hollow-

core FRP-concrete-steel bridge columns." Proc., Structures Congress 2015, 585-596.

Abdelkarim, O. I., Gheni, A., Anumolu, S., Wang, S., and ElGawady, M. (2015). "Hollow-

Core FRP-Concrete-Steel Bridge Columns Under Extreme Loading." No. cmr 15-008.

2015.

Abdulazeez, M. M., Abdelkarim, O. I., Gheni, A., ElGawady, M. A., and Sanders, G. (2017).

"Effects of Footing Connections of Precast Hollow-Core Composite Columns." Proc.,

Transportation Research Board (TRB) 96th Annual Meeting, Transportation Research

Board, Washington, DC.

Dawood, H., ElGawady, M., and Hewes, J. (2014). "Factors affecting the seismic behavior of

segmental precast bridge columns." Frontiers of Structural and Civil Engineering,

8(4), 388-398.

Teng, J., and Lam, L. (2004). "Behavior and modeling of fiber reinforced polymer-confined

concrete." Journal of structural engineering, 130(11), 1713-1723.

Teng, J., Yu, T., Wong, Y., and Dong, S. (2007). "Hybrid FRP–concrete–steel tubular columns:

concept and behavior." Construction and Building Materials, 21(4), 846-854.

Zhang, B., Teng, J., and Yu, T. (2012). "Behaviour of hybrid double-skin tubular columns

subjected to combined axial compression and cyclic lateral loading."